Retention — designs a lifecycle automation flow (onboarding / activation / winback / churn-save) as a trigger→message→timing→branch map anchored to one activation metric, scored by a quantitative critic. Capability metadata: routing.yaml. Methodology, philosophy, scope: references/playbook.md [PLAYBOOK].
Core question: "Does every message in this flow move the user toward ONE named activation metric, fire on a real product event, and stop sending the moment the user already did the thing?"
Critical Gates — read first
- Name the activation metric BEFORE writing any message. A lifecycle flow that isn't anchored to a measurable activation event ("created first project", "invited a teammate", "second session within 7 days") is a broadcast, not a flow. No metric → BLOCK and ask one question.
- Every send is event-triggered, never time-only-spray. Each step declares its trigger (product event OR elapsed-time-since-event) AND its suppression rule — the condition under which the user EXITS the flow (they already activated, unsubscribed, converted, or a higher-priority flow took over). A flow with no exit condition is spam.
- One flow, one job, one metric per artifact. Onboarding ≠ winback ≠ churn-save. Cross-purpose flows run the skill once per job. The flow-architect refuses to stack two activation metrics in one map.
- No PII, no list purchase, no consent bypass. The flow targets an OWNED, consented list. It never assumes a bought list, never writes a step that bypasses unsubscribe, never sends to non-consented addresses. This is a hard gate, not a critic dim.
- The flow is a design, not a send. This skill produces the flow map + per-step copy + measurement plan. It NEVER connects to an ESP or sends mail — publishing is human-owned + out of scope (see Guardrails in
references/playbook.md).
Quality Gate — 6 dimensions
Full rubric + Pass/Fail bands + per-dim scoring: references/rubric.md [PROCEDURE]. Critic agent: agents/critic.md.
- Gate: Total ≥30/42 AND every dim ≥4/7. Below either → FAIL, route to the named agent, max 2 rewrite cycles.
- Dimensions: Activation-fit · Trigger/Suppression soundness · Timing discipline · Branch logic · Per-step copy quality · Measurability.
Before Starting
Per references/_shared/before-starting-check.md. Recommended context:
| Artifact | Source | Required? |
|---|---|---|
research/product-context.md |
research-icp | Recommended — the product's "aha" + activation event + value moment |
research/icp-research.md |
research-icp | Recommended — persona, churn drivers, the job the user hired the product for |
brand/BRAND.md |
create-brand | Recommended — voice anchors + banned-language for the per-step copy |
docs/forsvn/artifacts/marketing/campaign-plan.md |
plan-campaign | Optional — the email/Mailbox channel's role in the broader GTM |
Pre-Dispatch + Mode
Run canonical Pre-Dispatch (references/_shared/pre-dispatch-protocol.md). Needed dimensions: flow-type (onboarding / activation / winback / churn-save) · activation-metric (the ONE event the flow drives) · entry-trigger (the product event or signup that starts the flow) · available product events (what the product can actually fire) · list-channel (email / push / in-app) · voice source. Missing activation-metric OR entry-trigger → hard-block, ask one question. Full Cold Start + Missing-Input Hard Blocks: references/procedures/pre-dispatch.md.
Mode (references/_shared/mode-resolver.md): budget: deep. --fast collapses to a single inline pass (one flow path, no branch enumeration, no rewrite loop). --fast does NOT skip Cold Start, Critical Gates 1-5, or the consent/PII hard gate.
Agents + Routes
4 sub-agents (flow-architect → copy → measurement → critic) across Layer 1 architecture → Layer 2 copy+measurement → critic gate. Two routes — A (compose a new flow), B (called by another skill, e.g. plan-campaign building the Mailbox channel). Full manifest + dispatch graph: references/agent-manifest.md [PROCEDURE].
Artifact Contract
- Path:
docs/forsvn/artifacts/marketing/design-lifecycle/[flow-type]-[date]-[slug].md - Lifecycle:
pipeline— re-run on activation-metric change, new product event, or churn-driver shift. - Frontmatter + body section order + the flow-map table schema:
references/format-conventions.md[PROCEDURE]. - Consumed by:
measure-results(the flow's per-step activation lift closes the loop),evaluate-content(per-step copy scoring).
Anti-Patterns
references/anti-patterns.md [ANTI-PATTERN]. Re-read before any flow ships: time-only spray, no suppression rule, vanity-metric anchor (opens not activation), every-step-same-CTA, winback-as-discount-reflex, onboarding longer than the activation window, ignoring the unsubscribe gradient.
Durable Rules (protected)
Completion Status
- DONE — flow map + per-step copy + measurement plan written, critic PASS.
- DONE_WITH_CONCERNS — flow complete but with a thin product-event inventory or an estimated activation baseline; flagged in the artifact.
- BLOCKED — flow-type and activation-metric imply different flows; needs a user scope decision.
- NEEDS_CONTEXT — no activation metric definable (product has no measurable value moment); recommend
research-icp/ product-context first.
Worked Example
End-to-end onboarding flow (B2B SaaS, activation = "first project shared with a teammate within 14 days", 5-step event-triggered map + 2 branches + suppression rules + measurement plan + critic PASS cycle 1, with a cycle-0 FAIL the critic caught): references/examples/lifecycle-walkthrough.md [EXAMPLE].
References
references/{playbook, rubric, format-conventions, anti-patterns, agent-manifest}.mdreferences/procedures/pre-dispatch.md[PROCEDURE]references/examples/lifecycle-walkthrough.md[EXAMPLE]agents/{flow-architect, copy, measurement, critic}.mdreferences/_shared/{pre-dispatch-protocol, before-starting-check, mode-resolver, marketing-foundations}.md